Toggle high contrast

Public Service Pensions: Consultation on the discount rate methodology

TUC response August 2021
Report type
Consultation response
Issue date
Key findings

In our submission to the consultation of the review of the discount rate in 2011, we concluded:

“The current approach to setting the discount rate used for SCAPE purposes based on the use of the STPR is correct. We are not aware of any coherent argument as for why HM Treasury should depart from a well established approach to making a current assessment of future expenditure. Nor have any arguments been advanced as for why a given financial commitment should be assessed different, compared to other commitments of a like amount, solely on the grounds that it relates to a public service pension scheme. Given this conclusion, there is no logical reason why the discount rate should not continue at the current rate of 3.5 per cent, unless and until it is changed for public spending more widely.”

We could add to this conclusion that the experience of the last 10 years has been that the use of a volatile forecast of GDP growth to set the discount rate has introduced instability, unnecessary costs, intergenerational unfairness and member detriment to the system. This has caused major disruption for employers, additional costs for government and importantly has hampered the delivery of public services. This disruption has raised major challenges for the members of TUC-affiliated unions who work hard to maintain high levels of service to the public.

Our strong recommendation is that to avoid a continuation of these challenges, we should return to the use of a discount rate methodology based on the Social Time Preference Rate. As well as being justified on a practical basis, there are strong points of principle which underpin our recommendation including the need for consistency in assessing public spending and the need to maintain intergenerational fairness.

Download full response (pdf)

Introduction

The TUC welcomes the opportunity to make a submission to Treasury on the discount rate to be used in setting the contribution rates for the unfunded public service pension schemes.

Our response to the specific questions raised in the consultation are set out below. It is helpful though to set out as an introduction four points which set the context for this debate and our contribution to it.

First, it is important to recognise that pensions are political and that decisions on the discount rate are not purely technical ones.  Public service pensions involve allocating resources between generations, between public and private sectors and across different workforces and communities. This supports the need for consistency between assessing spending on public service pensions and on other public projects which are evaluated using the principles in the Green Book.[1]

A key issue for members of defined benefit schemes is benefit security. Government is in a very different position to private sector employers in that it has a very strong covenant and can write promises in which members can have a high degree of confidence. There is no need for government to purchase additional security for these benefits as private sector employers do when they fund schemes at very high levels or invest the scheme assets in very secure, low yielding assets. Indeed, for the government to fund public service pensions on the basis that they should buy such protection would be an inefficient and unnecessary use of public funds. It is for this reason important to reject any claims that public service pension funding should be on a par with private sector funding. 

The TUC does not disagree with the argument that it is important to recognise the value of public service pensions for example in placing a value on the costs of employing public servants. But it is equally important to recognise the benefit of these schemes to employers, government and the taxpayer. Many of the roles performed by public servants are demanding and involve long periods of training, building up experience and institutional knowledge. It is important to the country as a whole that we attract the best people into these positions and that we then retain their knowledge and experience. The benefit of public service pensions is vital in helping public service employers to recruit, retain and encourage return of public servants and to help those employees to retire in part or fully. Without this tool, it would be far harder to manage the workforce to maximise service delivery. In addition to these clear commercial considerations, there are also wider benefits in having mutual schemes available across a service which can foster camaraderie and purpose for a service. As a society we should be proud of the high-quality pension benefits we provide to our public servants. It is for this reason that the TUC thinks that an additional objective for the review should be to ensure that good-quality public service pensions are affordable and available to support service delivery.

The TUC notes the parallel consultation being run on the cost control mechanism, but this evidence relates solely to the SCAPE discount rate. Our position is that the discount rate should not have any impact on the cost control mechanism.

Finally, we recognise that this consultation is an important part of a public debate on resource allocation. It is important that this is a wide and unconstrained debate. There are points in the consultation which hint at government wanting to narrow down the debate by focussing on the discount rate and not on any wider questions. Whilst we recognise the need to retain a clear focus, we do not think it is helpful to constrain the discussion too narrowly. In particular, one point we will cover in our response below is the issue of whether when discount rates change, there should be a rebasing of the notional assets of the unfunded schemes. It is important that government does engage with this issue which is a key part of the discount rate discussion rather than seeking to allocate it as outside the scope of the consultation.

Consultation questions

1.    Do you agree that these are the correct objectives for the SCAPE discount rate? If not, please explain why and specify any alternative objectives that you think should be included.

The TUC does not disagree with any of the three suggested objectives. In particular, the promotion of the objective of stability to being a key objective is right given the experience of the last decade.

Fair reflection of costs

The TUC agrees that a discount rate that generates a ‘fair’ reflection of costs is of paramount importance. In assessing fairness, the following issues should be considered:

  • It should be possible to make a fair comparison between spending on additional staff or other capital expenditure – which would imply consistency with the Green Book:
  • It should mean that the value placed on pension benefits for example by Pay Review Bodies is consistent with the costs to government of providing the benefit
  • It should mean that costs are allocated fairly across generations.

Reflect future risks to government income

The risks faced by a government in making future pension payments is a factor that should be taken into account in setting the SCAPE discount rate. However, the risk that the government will be unable to meet its commitment is relatively low, given its ability to levy taxation. The minimal risk which is faced is in no way specific to future pension payments as the money future governments have access to is ‘fungible’ – it could equally be used to pay off government debt, to fund road building or to pay public service pensions. In other words, the risk that at some point in the future the government would be unable to discharge its commitments would be shared equally by all those commitments.

This also supports the use of a discount rate that is consistent with that used to assess other government spending.

Stability

The experience of the past decade has been one where scheme costs have been unstable with the significant escalation in costs raising real challenges for employers including those private sector employers who provide public services. A volatile cost of pensions makes budgeting and workforce planning difficult for employers and this can have a negative effect both on employees and service delivery.

The instability has not been helpful for government who, as paragraph 3.10 of the consultation document points out, have had to find additional funding as a result of sudden, unexpected changes to the discount rate.

The TUC welcomes the promotion of stability as one of the key objectives in assessing different approaches to the discount rate methodology.

The consultation documents suggests that the only impact of the discount rate on the cost sharing mechanism is through any economic check that might be introduced. It is important to note that the discount rate does have a wider impact than this on cost sharing. This is because, with a lower discount rate the current 2 per cent cost corridor in cost sharing is a smaller proportion of overall costs. Breaches of the cost cap at the higher end are therefore more likely with lower discount rates and this element should not be overlooked in assessing stability.

Additional objective: encouraging high quality public service pensions

The TUC has consistently argued that a further objective in setting the discount rate should be to encourage and facilitate sustainable and affordable public service pensions that provide employees with an adequate pension and dignity in retirement.

As set out in the introduction, we do not argue for this objective on the basis of special pleading but on sound commercial reasons which support service delivery.

It is sometimes claimed that offering high-quality public service pensions is somehow unfair since these are not now widely available in the private sector. But it would not help those employees in the private sector defending their high-quality pensions to see public service pensions downgraded. Rather, high-quality public service pensions set a standard which the private sector needs to aspire to, particularly if the labour market tightens.

The fact that there are real problems that arise for public service employers when the value of public service pensions is undermined was demonstrated when the effect of annual allowance restrictions on doctors on the NHS Pension Scheme became clear. It is important to note that public service employers may not have many other levers to pull (such as share options and bonus schemes) in setting total reward packages. This all points to the importance of public service pensions and the need to facilitate their provision.

The TUC does not believe setting such as objective would conflict with wider government policy. When responding to the consultation on the McCloud remedy, the Chief Secretary to the Treasury noted that “Public service workers provide vital services that all of us count on, and their unwavering commitment is inspiring, particularly, as we face down the coronavirus pandemic. It is a long-standing practice that the overall reward package for public servants includes a generous pensions element.”[2]

Government pensions policy in general also recognises the importance for workers and the wider economy of maintaining good quality occupational pensions. Speaking at the second reading of the Pension Schemes Bill in 2020, secretary of state for Work and pensions Therese Coffey said: “This unprecedented period that we have been experiencing has shown more than ever the need for financial resilience but also the need to focus on future resilience. Helping workers to achieve greater financial resilience for themselves for the long term is a crucial part of our economic recovery. Improving the financial resilience of the public is a personal priority for me and I am proud that the Bill is designed to help pension showing location of savers across the country.”[3]

Meeting these policy objectives would be made harder with a lower discount rate used in the SCAPE methodology.

2.    Do you agree that these are the most appropriate methodologies that should be considered? If not, please specify any alternative methodologies that should be considered and how they would fit with the government’s proposed objectives.

As stated earlier, the TUC is anxious to ensure that the debate on these important issues is not constrained. In our response below and in particular in our response to question 3, we include as part of the methodology the principle that if the discount rate changes, this should be reflected in a rebasing of the notional assets. We set out the principles of this part of the methodology here.

It is now very common in the private sector for pension schemes to ‘hedge’ themselves against changes to the bond returns that tend to drive the discount rates used in their valuations. To use a very simplified example, a scheme might hold a portfolio of bonds whose coupon and capital payments matched the scheme’s benefit outflow. The scheme might be valued using a discount rate of say 3 per cent and assume that on that basis the assets and liabilities had the same value. If in the market, the yield on bonds increased, this would cause the value of the bond portfolio to fall as investors would pay less for the fixed return the bonds paid. But if the scheme were then revalued, the discount rate would also rise reflecting the higher rates of return available in the market. In this simple example where the cash flows match, the effect of the yield change on the assets and liabilities would be the same and the so the yield change would not cause any deficit or surplus in the scheme. In other words the scheme is immunised against changes in bond yields.

In reality, schemes need to use more complicated structure and financial instruments to achieve this matching of assets and liabilities, but the principles remain the same. These approaches are often referred to as Liability Driven Investment or LDI. Schemes have adopted these approaches as scheme trustees and employers have recognised that without hedging, changes in bond yields can cause considerable instability in funding levels and in the cash calls on employers.

Note that with an LDI approach, the scheme is not protected against other factors that might affect the liabilities – surpluses or deficits arising from changes in longevity or retirement patterns or pay increases will still come through.

The TUC believes that whatever methodology is adopted, it should be coupled with a principle that on any change in the discount rate, the notional assets will be rebased to remove the effect of the discount rate change on accrued liabilities. Note that the proposal is not to rebase the assets to set them equal to the new liabilities. It is to rebase them purely to allow for the discount rate change.

Assessing this aspect of the methodology against the objectives for the review, the TUC would make the following comments:

·       This approach would be fairer in particular in avoiding the current approach where employers are required to pay now and in the future for liabilities that arose over the past decades – perhaps when they did not exist as an employer in the scheme. To the extent that additional funding from employers has an impact on the resources available for employees, the current position involves significant intergenerational transfers. For example the effect of the falling discount rate used over the past decade where the SCAPE discount rate has been set in line with GDPR forecasts has created significant deficits in schemes. These deficits are being met over the coming decade by employers whose employees include a new generation of workers who did not contribute to the deficit. Note that these issues would also arise with an increase in discount rates – which would create a surplus to be shared over the “wrong” employers and employees.

·       It might be argued that this approach is unfair in that it does not charge employers for past underpayments. A counter to this would be that past payments were thought to be sufficient at the time and, if real assets had been involved, could have been invested at the yields then available.

·       This approach is also fair in that it does not remove past experience in areas such as mortality, commutation, family statistics, retirement patterns, from costs that do need to be shared, It does retain the characteristic of “adjusted for past experience”

·       This approach does not have any effect on the risks to future government spending

·       This approach would significantly increase the stability of any approach that involved changes to the discount rate from time to time. The effect has varied from scheme to scheme but the impact of past service deficits has been a significant contributor to the volatility of costs over the past decade.

·       The approach would have also addressed in part the outcome of the cost sharing valuations which indicated a reduction in the value of accruing benefits and an increase in employer contributions. It was surprising that in their analysis of the cost sharing mechanism, the government Actuary concluded that this outcome was “peverse” rather than analyse how the discount rate methodology had affected the outcome

3.    What are the advantages and disadvantages of a SCAPE discount rate methodology based on expected long-term GDP? If this methodology is adopted, should any of the modifications (allowing for short-term GDP projections, allowing for actual experience) be considered?

The TUC has consistently argued that expected GDP growth is not a suitable basis on which to cost public service pensions. The experience over the last decade has strengthened our belief that this approach is not helpful when judged against the objectives set for this review.

An advantage claimed for this approach is that it establishes a relationship between future tax revenue and pension scheme benefits and therefore provides a check on the affordability of public service pensions. It is hard to justify why this is an affordability check that should apply to public service pensions but not to other elements of government spending which are appraised using the methodology set out in The Green book and in particular use a social time preference rate for discounting.

On the fairness and stability objectives, this methodology has not performed well. As the consultation document points out, there have been three changes to the discount rate used under this approach since 2011. This means that employers have paid different costs for the benefits accruing depending on the time at which these benefits were accrued. It is hard to argue that those different costs have been fair or that they represent a pension which has a fundamentally different value to either the employer or employee based on when the accrual occurred over the last ten years.

As an example of unfairness, in 2019, employer contributions to the Teachers’ Pension Scheme increased from 16.48per cent to 23.68per cent - an increase of over 40per cent. A large portion of this additional cost was due to discount rate changes. This resulted in many independent schools leaving the scheme and moving teachers often to far inferior schemes such as minimum auto-enrolment schemes. So a teacher working in 2017 could receive a years accrual of good quality defined benefit pension whereas a teacher working in 2020 would receive a low level contribution to a defined contribution scheme. These are the real world impacts caused by the current methodology.

The instability of the methodology has also made it difficult for employers to budget for or understand their pension costs and this will have affected service delivery. As noted in the previous question, the instability in this approach could be tempered by moving to a methodology where notional assets are rebased in the event of a change in the discount rate.

As we do not support the use of a GDP based methodology, we do not comment on the potential modifications to it.

If government were to retain a GDP-based methodology, this should be coupled with a methodology which requires that assets are rebased on the change in discount rate as explained in question 2 above.

4.    What are the advantages and disadvantages of a SCAPE discount rate methodology based on the STPR? If this methodology was adopted, should any modifications (allowing for the public service pension context or allowing for long-term uncertainties) be considered?

The TUC has been consistent in its support for the use of a social time preference rate (STPR) approach to the discount rate used for valuing public service pensions. n

The principle behind our support for this method is that it uses a consistent approach across all financial decisions taken by the government at a point in time that is removed from when the decision is given effect. Much of the criticism of unfunded public service pensions arises from a failure to understand the nature of unfunded pension liabilities in the public sector and how, more generally, political judgments should be made on issues of public policy that depend on income or expenditure that arises at some time in the future. These decisions involve inter and intra generational judgements. There are wide areas of policy where these judgements have to be made and it is important that consistent evaluations are made to weigh up the costs of different elements of government spending. (Assume for example that government wishes to compare the costs of employing a large number of data analysts compared to the cost of investing in building an Artificial Intelligence system. The costs of the data analysts include their pension costs and these need to be discounted using the same methodology to produce a fair comparison).

The approach used in making these assessments has been subject to significant academic and political debate and the current consensus is set out in the Green Book.

Green Book Guidance

The Green book is very clear in stating that:

“Green Book guidance applies to all proposals that concern public spending, taxation, changes to regulations, and changes to the use of existing public assets and resources. Green Book guidance covers:

·       policy and programme development

·       all proposals concerning public spending

·       legislative or regulatory proposals

·       sale or use of existing government assets – including financial assets

·       appraisal of a portfolio of programmes and projects

·       structural changes in government organisations

·       taxation and benefit proposals

·       significant public procurement proposals

·       major projects”

It is hard to see how spending on public service pensions falls outside this scope. It is also noticeable that the guidance applies to benefit proposals – and again it is difficult to see why spending on old age benefits should be different to spending on public service pensions. The current approach which values public service pensions on a lower discount rate would suggest that it is better for government to provide State pension benefits for public service employees rather than occupational pensions.

On the STPR, the Green Book says in para 5.33:

“For individuals, time preference can be measured by the real interest rate on money lent or borrowed. Amongst other investments, people invest at fixed, low risk rates, hoping to receive more in the future to compensate for the deferral of consumption now. These real rates of return give some indication of their individual pure time preference rate. Society as a whole, also prefers to receive goods and services sooner rather than later. This is known as ‘social time preference’. The discount rate used in the Green Book is known as the ‘social time preference rate’ (STPR). It is the rate at which society values the present compared to the future”

In essence, the Green Book makes it clear is that when policy decisions have to be made now that involve income or expenditure in the future, there is a proper distinction to be made between the discount rate that it would be appropriate for individuals to use and the discount rate that should be used for society as a whole. Put simply, there are things that society as a whole can do that are just not open to individuals. One of the key differences is that the individuals are subject to the exigencies of investment markets, while society as a whole is not. As a result the STPR does not go up and down with movements in markets, but is set for the long term.

There is wide and deep selection of reading on the Social Time Preference rate – sometimes also called the Social Rate of Time Preference – both nationally and internationally. It is generally viewed as being made up of two elements – the pure time preference which would apply if economies did not grow and a wealth effect which allows for the fact that with growing economies, a fixed spend accounts for a smaller part of a future society’s “pie”.

The Green Book sets the STPR at 3.5per cent, a rate that has been fixed for a considerable time.

International comparisons indicates that a STPRs in the range 3per cent to 4per cent are used widely across a number of economies.[4]

Assessment Against Objectives

The use of a STPR approach provides the best fit to the policy objectives set by government:

the approach is fair in providing a consistent approach to costing alternative government spending.

In particular the approach will remove much of the intergenerational unfairness in the current approach Using a stable STPR rate to assess costs would avoid the intergenerational transfers that have been a feature of the mechanism used over the past decade

Similarly using the STPR approach also allows government to assess the risks from public service pensions on a consistent basis with other risks

It provides a stable methodology which will remove the volatilities which have caused significant challenges for the public services and their employees over the last decade.

Modifications to the STPR

The consultation document suggests that one modification which might be applied to the STPR is to remove the catastrophic risk element. The consultation document provides no rationale for this, simply stating that “it might be considered inappropriate”. The perhaps unhelpfully named “catastrophic risk” element is meant to cover unforeseen changes in social and political objectives and wider changes in the economy, society and technology. It is difficult to see any reason why public service pensions would not be subject to these uncertainties. We would therefore argue strongly against the removal of this element. It is also worth commenting that a key advantage of using the STPR is consistency in the treatment of spending. Removing this element would undermine that consistency.

This consistency principle would support the use of lower discount rates to value liabilities that fall in the long-term future and the consultation suggests a lower discount rate of 3per cent beyond 30 years and 2.5per cent beyond 75 years. It would be helpful to know whether these lower discount rates are actually used in practice in appraising other projects. It would also be helpful to understand the effect of using these rates in practice on scheme costs. Finally, consideration should be given to the additional complexity this might introduce into valuations and whether that would be justified given the small impact this might have on overall results.

5.    Which SCAPE discount rate methodology do you recommend, and why?

The TUC strongly recommends the use of the social time preference rate. This approach best meets the objectives set by Treasury for the review and also meets the additional objective we have specified.

6.    Are there any equalities impacts of changes to the SCAPE discount rate methodology that the government should consider?

It will be important to conduct a full equality impact once a decision has been taken on the appropriate methodology. Given the area covered here, part of that assessment should be to consider intergenerational fairness issues. 

7.    Do you agree with the proposal for reviews of the SCAPE discount rate to be aligned with the scheme valuation cycle?

The TUC strongly supports the aligning to the review of the SCAPE discount rate reviews with the scheme valuation cycles.

 

 


[1] The Green Book: appraisal and evaluation in central government -  www.gov.uk/government/publications/the-green-book-appraisal-and-evaluation-in-central-governent

 

[2] Public service pension schemes: changes to the transitional arrangements to the 2015 Schemes - Government response - Public_Sector_Pensions_Consultation_Response.pdf (publishing.service.gov.uk)

[4] See Public Sector Discount Rates: A Comparison of Alternative Approaches, John Creedy and Hemant Passi  New Zealand Treasury Working Paper 17/02 2017

 

Enable Two-Factor Authentication

To access the admin area, you will need to setup two-factor authentication (TFA).

Setup now